[Cschwatz]

Hello CFD wizards,

I've been trying for the past days to develop an UDF that provides a pulsatile behaviour for an inlet flow. This pulsatile behaviour tries to simulate the same behaviour seen in a heartbeat, which is a flow rate/velocity that varies over time due to contractions and relaxations of the heart over a cardiac cycle.
Basically, what I've done is a regression of the data that I've obtained from an article and separated them in a few polynomial functions, since it would've been quite complicated to get the pulsatile behaviour in just a single polynomial function.

When I compile the UDF, the velocity presents a constant behaviour, not following what is expected.
Also, is there any way that I could make a loop for this UDF, so I just update the time and pull a specific function for that time? Given that these functions will repeat themselves every cardiac cycle.

Thanks in advance!

The UDF that I have, so far, is as follows:

#include "udf.h"
DEFINE_PROFILE(unsteady_velocity, thread, position)
{
face_t f;
real t = CURRENT_TIMESTEP;

begin_f_loop(f, thread)
{


if (t>=0.0 && t<= 0.128)
{
F_PROFILE(f, thread, position) = -122.25*pow(t, 3) + 25.689*pow(t, 2) - 0.1683*t + 0.0701;
}

else if (t> 0.128 && t<= 0.186)

{
F_PROFILE(f, thread, position) = 170.88*pow(t, 3) - 74.116*pow(t, 2) + 10.038*t - 0.2159;
}

else if (t> 0.186 && t<0.261)

{
F_PROFILE(f, thread, position) = -10.052*pow(t, 3) + 5.9452*pow(t, 2) - 0.09495*t + 0.2198;
}

else if (t >= 0.261 && t<0.388)

{
F_PROFILE(f, thread, position) = 41.661*pow(t, 3) - 43.059*pow(t, 2) + 14.028*t - 1.2703;
}

else if (t >= 0.388 && t < 0.468)

{
F_PROFILE(f, thread, position) = -59.067*pow(t, 3) + 77.805*pow(t, 2) - 33.944*t + 5.0295;
}

else if (t >= 0.468 && t < 0.687)

{
F_PROFILE(f, thread, position) = 6.8298*pow(t, 3) - 12.125*pow(t, 2) + 6.8967*t - 1.1411;
}

else if (t>=0.687 && t<1.0)

{
F_PROFILE(f, thread, position) = 1.8829*pow(t, 3) - 5.1168*pow(t, 2) + 4.523*t - 1.2208;
}

}
end_f_loop(f, thread)
}

[AlexanderZ]

#include "udf.h"
DEFINE_PROFILE(unsteady_velocity, thread, position)
{
face_t f;
real t;
t = RP_Get_Real("flow-time");

begin_f_loop(f, thread)
{

t =  fmod(t, 1.0);

if (t>=0.0 && t<= 0.128)
{	
F_PROFILE(f, thread, position) = -122.25*pow(t, 3) + 25.689*pow(t, 2) - 0.1683*t + 0.0701;
}

else if (t> 0.128 && t<= 0.186)

{	
F_PROFILE(f, thread, position) = 170.88*pow(t, 3) - 74.116*pow(t, 2) + 10.038*t - 0.2159;
}

else if (t> 0.186 && t<0.261)

{ 
F_PROFILE(f, thread, position) = -10.052*pow(t, 3) + 5.9452*pow(t, 2) - 0.09495*t + 0.2198;
}

else if (t >= 0.261 && t<0.388)

{ 
F_PROFILE(f, thread, position) = 41.661*pow(t, 3) - 43.059*pow(t, 2) + 14.028*t - 1.2703;
}

else if (t >= 0.388 && t < 0.468)

{
F_PROFILE(f, thread, position) = -59.067*pow(t, 3) + 77.805*pow(t, 2) - 33.944*t + 5.0295;
}

else if (t >= 0.468 && t < 0.687)

{
F_PROFILE(f, thread, position) = 6.8298*pow(t, 3) - 12.125*pow(t, 2) + 6.8967*t - 1.1411;
}

else if (t>=0.687 && t<1.0)

{ 
F_PROFILE(f, thread, position) = 1.8829*pow(t, 3) - 5.1168*pow(t, 2) + 4.523*t - 1.2208;
}

}
end_f_loop(f, thread)
}

best regards

[Cschwatz]

Thanks for the reply Alexander!

However, I'm still encountering the same problem as before. When I run a simulation with this udf, the velocity remains constant throughout the entire simulation. For some reason, the functions that I've written in the code are "not being used" by fluent. Is there a way to check where the problem could be in the code?

Thanks again!

[AlexanderZ]

code is correct

did you apply this profile to your boundary?

best regards

[Cschwatz]

Yes, I used it as an inlet boundary condition. However, for some odd reason, when I plot the velocity profile via fluent monitors, it just plots a straight line, like if it was ignoring all the polinomial functions I've written in the code. I've also tried to run the UDF by compiling it and by interpreting it, but there was no difference.

Thanks once again!

[AlexanderZ]

compile code I've sent before.
I've applied it as temperature BC and it works, it should works as velocity BC as well.

best regards

[Cschwatz]

I've compiled the code as you said and I still got the same problem...
I've attached 2 screenshots of the fluent window to show you the output the udf is giving. As you can see, instead of plotting a profile (as expected), it just plots a straight line, as if the velocity was constant throughout the entire simulation. Not sure if I'm doing something wrong in the C# script, since I just copied and pasted the code you've sent me before.

Thanks again


https://imgur.com/a/bUtXjxq